WRIGHT BROTHERS — BAKER 213 



They made many successful glides with the machine as altered, 

 but upon trying it as a kite, as originally intended, they found that 

 the lift was less than one-third of their predictions. To test whether 

 this was caused by porosity of the wing covering, two small test sur- 

 faces were measured in natural wind with negligible difference. 



Aside from the valuable flight experience gained with this machine, 

 the summer's observation taught them the following: 



1. Published lift characteristics for curved surfaces were definitely 

 in error. 



2. Over-all efficiency depended upon L/D rather than lift alone. 



3. The relative position of the upper and lower wings decreased 

 tlio theoretical total lift of the individual surfaces, that is, they noticed 

 biplane effect. 



4. The customary method of expressing the air force acting on a 

 wing in terms of a pressure normal to the chord line had led them 

 into a misconception of the net lift and drag components. Although 

 they did not express their understanding of this in these words, there 

 seems little doubt that this was the primai-y motive in designing the 

 test instruments that will be described later. 



In a paper read September 18, 1901, before the Western Society of 

 Engineers, primarily a society of civil engineers, Wilbur Wright men- 

 tioned a series of experiments that they had begun for measuring 

 the magnitude and direction of the forces acting on various types of 

 curved surfaces. We know now that these experiments began upon 

 their return from North Carolina and followed a rather interesting 

 development pattern. 



The first attempt at measuring the characteristics of a surface in 

 model size is shown in plate 2, figure 1. Here we see a bicycle wheel 

 mounted horizontally on tubes extending forward from the handle- 

 bars of an ordinary bicycle (of their own manufacture). At one 

 point on the wheel was mounted a flat plate. Some 120° removed 

 from this point was mounted the model surface. The test surface 

 was adjusted in angle of attack until its lift would just balance the 

 flat-plate resistance riding, normal to the wind when the bicycle was 

 pedaled forward. 



In a letter to Octave Chanute, October 6, 1901, Wilbur told about 

 using this device and how they were able to check the ratio of the 

 lift of a surface at any angle of attack versus its flat-plate resistance. 

 Also, he noted that the Smeaton formula for flat-plate resistance, 

 P=0.005A72^ as used by the United States Weather Bureau, was 

 evidently in error, and suggested that a constant of 0.0033 would be 

 nearer the truth. 



In this same letter, Wilbur stated that the bicycle test was very poor 

 for measuring surfaces at small angles and went on to describe their 



